Different speed tests give different results

[codie]

I have read the various threads on this topic already posted over a number of years, including your original pinned "Why Do My Results Differ From Speedtest.net / Ookla Speed Tests?" I attach a few sequential tests supporting your assertion! My query tho is this: over the past year I have had ongoing problems with my ISP (TalkTalk in the UK - it heads the league in number of complaints from customers saying their broadband speeds are atrocious and nothing like what are promised!). Anyway, for the past 4 weeks I have actually engaged with their fault reporting system, even tho it is a major pain (waiting in for engineer visits/holding phones which never get answered/repeating the problem & diagnostics for the 100th time etc etc) - Consequently I have been using your site (auto up/download every 5 mins 200x) & occasionally comparing the speeds you report to the ones I get on eg Ookla measured a few minutes later. Until my a few days back, I have actually found good agreement between the two sets of measurements - a mean download speed of around 40Mbps on both sites, and successive measures on the 2 sites not differing by more than 2-5 Mbps (so your site gives a reading of eg 39.2Mbps, & the speedtest.net a few minutes later giving a reading of eg 42.9Mbps). Basically differences I can live with. However, following yet another engineers visit (I should be getting closer to 76Mbps, but of course they say they will provide between 48 & 70 Mbps) I found the two speed test sites diverging - you are consistently showing < 40Mbps (and fluctuating (downwards)), whilst they show steady download speeds of 50 - 53 Mbps - this of course taking them above their minimal agreed provisioning rate, but as far as I am concerned is capped. I am also dubious of the consistency in their repeated measures. Unfortunately, what I have written above represents about the limit of my technical understanding, & so I will just have to live with it. I wrote a formal letter of complaint yesterday with a load of technical data backing up my argument (so THANKS Testmy.net!!), but other than threatening naming and shaming to OFCOM there is little to do other than withdraw my custom, but the other ISPs are probably little better! What amazes me is that delivering an acceptable broadband service to customers in London should not be rocket science - what is the problem?!!